What Can Prediction of Emotional Regulation Teach Us?

We examined the performance of regression- and machine-learning (ML) algorithms in predicting college students’ (N= 1,077) adaptive and maladaptive emotion regulation use. Regression and ML performed comparably well (MSE = 0.011 - .0141, R-squared = 45– 50), though ML retained more predictors (features) and explained more variance.